Sixty-five arbitrage windows opened today, led by a 20.43% Gate Futures-to-KuCoin gap on EVAA, which alone accounted for five of the ten widest spreads. Real edge exists in this dataset, but thin books, futures-to-spot bridging costs, and missing depth data mean most of these won't survive contact with real size.
Today's scan turned up 65 discrete arbitrage windows across the venues we track — Binance, Binance Futures, Bitget, Bitunix, Coinbase, Gate Futures, and KuCoin all showed up on one side of a spread or the other at some point. The headline number is EVAA's 20.43% gap between Gate Futures at $0.521000 and KuCoin at $0.542104, the widest print of the day and one of five separate EVAA dislocations that made the top ten alone. Second place goes to APE at 14.59% — Coinbase $0.135000 against Binance $0.154700 — a geography-flavored gap shape that's become familiar this cycle.
Set expectations before scrolling further: none of these numbers come with a depth snapshot or a duration stamp in the feed. What we have is price and percentage, not order book size or how long the window stayed open, so treat everything below as "here's where the inefficiency showed up," not "here's a guaranteed fill." Double-digit spreads on lower-float names are as often a stale quote or a thin ask as they are real, executable edge. The job today is sorting which of these 65 are worth a bot cycle and which are noise.
Buy side: Gate Futures at $0.521000. Sell side: KuCoin spot at $0.542104. A double-digit gap between a futures mark and a spot ask is the classic shape of a basis dislocation rather than a clean cash-and-carry, and it matters operationally: you can't withdraw a futures position straight to a spot exchange. You have to close the future, take proceeds in stablecoin, buy EVAA spot somewhere, then move it to KuCoin to sell — three hops, each adding latency and slippage risk. On a token trading at roughly half a cent, a few seconds between legs can eat a meaningful slice of a 20% headline. Executable for a bot with capital pre-funded and idle on both venues; not realistically executable as a manual trade unless you already hold EVAA on the Gate side and stables on KuCoin before the window opens.
Buy at Coinbase $0.135000, sell at Binance $0.154700. This is the more conventional arb shape — a retail-heavy US exchange lagging the deeper, faster Binance book. APE has enough trading history that this isn't a listing-day fluke; it reads as a real but short-lived lag during a fast move, the kind that opens for seconds to low minutes before Coinbase's book catches up. Bridging an ERC-20/L2 balance from Coinbase to Binance adds gas plus a confirmation wait that can run anywhere from under a minute to 10+ minutes depending on network congestion — long enough for a chunk of that 14.59% to close before the tokens land. Executable for a bot with capital pre-positioned on both sides; a manual trader chasing this after the fact is likely buying the top of an already-closing gap.
Buy $0.885200 on KuCoin, sell $0.926400 on Bitget. Note the price here is roughly 70% higher than the Gate Futures/KuCoin print in #1, meaning EVAA moved hard between these two snapshots and stayed volatile enough to throw off multiple simultaneous spreads across different venue pairs on the same day. That's a signal on its own: a token appearing five times in one day's top ten spreads is either freshly listed on one of these venues, thinly traded, or both. Good news for arb hunters in that the inefficiency is real and repeating; bad news in that repeated large spreads on the same name usually mean at least one venue's book is too shallow to trust the quoted price for real size.
Buy $0.123900 on Binance Futures, sell $0.137700 on Bitget spot. Same futures-to-spot structure as the EVAA #1 trade, with the same caveat: closing a futures position and routing proceeds to a spot buy isn't a single atomic hop. B shows up twice in today's top ten (also #10 further down the list), both times with Bitget on the sell side — worth flagging as a pattern rather than a coincidence, covered in the next section.
Buy $0.473044 on Bitget, sell $0.520900 on Binance Futures — the mirror image of the B trade above, with Bitget on the buy side this time instead of the sell side. Double-digit spreads on lower-cap names like LAB aren't shocking, but this is also the smallest of the top five, which means fee and slippage drag eats a proportionally bigger share of the edge than it does on the EVAA or APE prints.
Two shapes dominate today's top ten. First: futures-vs-spot basis gaps. Gate Futures or Binance Futures sits on one side of five of the ten listed spreads — as the cheap buy leg against spot four times (EVAA #1 vs KuCoin, B #4 vs Bitget, EVAA at 9.06% vs Bitunix, EVAA at 8.33% vs Bitunix) and, once, flipped to the expensive sell leg instead (LAB #5, Bitget buy vs Binance Futures sell). That flip matters: the futures mark isn't drifting in one consistent direction relative to spot, it's tracking wherever that specific token's perpetual is thin enough to diverge that day. Treat each futures-spot gap as its own read, not a standing directional bias.
Second: Bitget is the single busiest venue in this list, appearing in five of the ten spreads — selling into two EVAA prints and one B print, buying into LAB and the second B print. Whether that reflects Bitget consistently running a step behind on price discovery for smaller-cap names, or simply being first-to-list them, it's the venue worth a standing quote feed if you're building a monitor. Coinbase and Binance paired up twice, and in opposite directions on the same day — APE went buy-Coinbase/sell-Binance, while T went buy-Binance/sell-Coinbase. Two spreads on the same exchange pair running opposite ways in one session is a sign US-session volatility was actively whipsawing the Coinbase-Binance basis rather than sitting in one steady lag direction.
The tradeoff here is straightforward and unforgiving. A 20% spread on a token like EVAA looks enormous, but if the book is only three levels deep before price moves against you, a $500 probe clip fills clean while a $10,000 clip walks straight through the spread and hands half the edge back as slippage. The bigger the headline percentage, the more likely it's sitting on a thin book — that's not coincidence, it's cause and effect: thin books are exactly what let big spreads form and persist in the first place. Smaller, faster spreads — T at 9.67%, B at 8.31% — on names with more established trading history are more likely to hold their quoted price through a real-size fill than the more dramatic EVAA prints.
Practical sizing: start with a probe clip sized at 5-10% of what you'd normally run on a liquid pair, confirm the actual fill price matches the quote within a few basis points, and only scale up on the second or third pass. If the probe clip already moves price more than 1-2%, the headline spread was never really there for size — it was a top-of-book artifact, not tradeable depth.
Two worked examples, with fee assumptions stated explicitly since the feed doesn't include a live fee schedule.
EVAA is the name to keep a standing alert on — five top-ten appearances in one day usually means the underlying liquidity conditions, whether a new listing, low float, or both, aren't resolved overnight. Expect similar gaps tomorrow, especially around Gate Futures paired against KuCoin or Bitget. B and LAB's Bitget-centric pattern is worth a standing monitor too; if Bitget keeps sitting on one side of these smaller-cap spreads, it's the venue to check first when a new gap opens. On the majors side, watch Coinbase against Binance during any fast US-session move — that pair produced both the APE and T spreads today, in opposite directions, which suggests it's actively whipsawing rather than settling into a steady lag. Best general watch windows: the first 30-60 minutes after Asia-session open and the first 30-60 minutes of US cash equities open (roughly 9:30am ET), when liquidity handoffs between regional order books are most likely to leave a gap open long enough to hit.
Sixty-five windows, one big headline number, and a lot of fine print underneath it. Chase the ones with real depth behind them, skip the ones that only look good on paper, and don't forget that a spread you can screenshot isn't the same as a spread you can fill. Arbitrage Hunter — July 12, 2026.